Spectral Evaluation of Hyperspectral Remote Sensing Data Using UAV-derived VIS/NIR/SWIR Imaging Data and Polygon Grids

In 2022, the hyperspectral mission EnMap was successfully brought into orbit with some delay, but is currently the only remote sensing mission providing hyperspectral image data in 30 m spatial resolution. More hyperspectral satellites with a similar spatial resolution are expected. Planet has announced the launch of two hyperspectral satellites for 2023. Without doubt, a new hyperspectral remote sensing area began asking for validation and calibration workflows. In this context, we propose to focus within the full range between 350 nm and 2,500 nm on selected 18 bands. These bands represent characteristic reflectance and absorption feature, especially of vegetation. In this study, we present the proposed workflow for one band (910 nm) which was obtained from a newly developed VIS/NIR/SWIR imaging system for UAVs. The UAV-acquired NIR data has a very high resolution (< 0.1 m) and was downscaled to 30 m using polygon grids and zonal statistics. The UAV data was then plotted against downscaled Planet SuperDoveSat and SkySat NIR bands. The SkySat data correspond well (R2 = 0,72) with the UAV NIR data while the Planet SuperDoveSat data performed moderate (0.57).